Tree rings are a lousy proxy for temperature. John Daly explains:
"Tree rings are the primary proxy behind the `Hockey Stick', particularly the earlier part of the millennium. Tree rings are only laid during the growing season, not the whole year, and so they tell us little or nothing about annual climate. For example, this year (2000) there was a warm winter and early spring in the north-eastern USA, followed by an unusually cool summer and fall. Since the two events are largely self-cancelling, the year may finish as fairly average, but the tree rings would only record the cool summer and thus give a completely false impression of the full-year temperature. Tree rings do not even record night temperatures since photosynthesis only occurs in the daytime. Yet winter and night temperatures are an essential component of what we understand by the concept `annual mean temperature'.
All a tree ring can tell us is whether the combined micro-environmental conditions during the growing season were favourable to tree growth or not. This is because tree rings are influenced by numerous factors other than temperature, such as rainfall, sunlight, cloudiness, pests, competition, forest fires, soil nutrients, frosts and snow duration. Thus they are not even a good daytime temperature proxy for the few months of the growing season. Other proxies such as isotopes in coral, ice, minerals and sediments are vastly superior.
Trees only grow on land. Since 71% of the planet is covered by oceans, seas and lakes, tree rings can tell us nothing about the maritime climate, even though the oceans are known to be the prime determinants of climate conditions throughout the world.
In other words, historical climate simply cannot be described without taking into account the winter and adjacent months temperatures, night-time temperatures, and ocean sea surface temperatures. Tree rings, no matter how carefully they are measured and examined, cannot provide information on any of these key parameters, and are a doubtful proxy even for daytime temperatures on land in summer.
A final weakness arises when calibrating the tree rings against temperature. When measuring the width or density of a tree ring, exactly what temperature is represented by that measurement? This can only be determined by calibrating recently laid rings against known temperatures that existed at the time. Even this is problematic as the `known temperatures' can mean using a temperature series seriously contaminated by heat island and other local errors. If the calibrating temperatures are wrong, the whole tree ring temperature reconstruction for the distant past is also compromised.
There are many sub-specialties within the greenhouse sciences, `dendrochronology' (study of tree rings) being one of them. That particular sub-branch has both prospered and been highly successful in projecting itself to the broader climatic community on the basis of what is a very weak proxy.
In respect of Europe and Greenland, the IPCC and `National Assessment' do not challenge the existence of the Medieval Warm Period and Little Ice Age as they are too well recorded in other proxy indicators and historical accounts of the time. Instead, these events are now presented as being purely local to Europe and Greenland, but completely absent elsewhere in the world.
In general, the greenhouse industry disregards historical evidence, claiming them to be merely `anecdotes'. However, the idea that historical evidence can be easily dismissed as `anecdotes' in favor of questionable proxies like tree rings is to suggest that professional historians cannot be trusted to be objective.
Objectivity comes from how the evidence is treated, not the nature of the evidence itself. Historians can be just as objective as any scientist. Indeed most of them regard their work as science. As a prominent Finnish scientist remarked about a historical military event in his country's distant history, "if `anecdotal' ice is thick enough to carry a whole army, we can infer the ice was both thick and durable as an objective conclusion based on a documented historical fact."
Similar inferences can be made elsewhere in the world. For example, if whole populations suffered from drought-induced famine, we can infer a reduced rainfall. We don't need the proxies to tell us - indeed they might even mislead us. When a society is ravaged by great floods, we can infer increased precipitation. When the Polynesians were able to populate the Pacific Islands by outrigger boats, we can make climate inferences there too.
The fact that the greenhouse sciences were reluctant to declare the Medieval Warm Period and Little Ice Age as non-events in Europe suggests that the historical evidence was too overwhelming to make selected proxies believable. Such a claim for Europe would have been met with derision. While greenhouse science may regard proxies as being more objective than historical `anecdotes', that viewpoint is only shared among that peer group. The wider academic community, governments, and public opinion (the most important peer group of all) will give much more credibility to well-researched historical evidence.
If the IPCC were genuine about the need for full information about millennial climate, they would involve historians everywhere to research their resources to determine past climates as observed and experienced by human societies. The fear of some global warming proponents is that the historians would indeed find the Medieval Warm Period and Little Ice Age all over the world and that governments and public opinion would accept the historical accounts over tree rings."
http://www.john-daly.com/hockey/hockey.htm